1. Introduction
Internal waves are nearly ubiquitous in the oceans (Garrett & Munk, 1979), carrying energy vertically from the surface downward and from the sea-floor upward, as well as horizontally, in some instances over ocean basin-scale distances. Internal wave breaking ultimately dissipates much of the wind and tidal energy in the oceans (Waterhouse et al., 2014; Kunze, 2017), and deep internal wave mixing drives water-mass transformations important to the global deep ocean circulation, sometimes called the meridional overturning circulation (Munk & Wunsch, 1998; de Lavergne et al., 2016; Kunze, 2017; Nikurashin & Ferrari, 2013). Internal waves are often measured either from time series of isotherms (analyzing their vertical displacements or vertical strain) and horizontal velocity (analyzing the velocities or vertical shear of those velocities) (e.g., Alford et al., 2016). Here we use engineering data from the descent portion of Deep Argo float cycles to assess the amplitude of vertical velocities and vertical wavelengths of internal waves in the water column below the thermocline.
Deep Argo is a relatively new mission for the Argo program, designed to regularly sample the global ocean below the 2000-dbar pressure limit of core Argo floats, and hence improve monitoring of variations in ocean temperature, salinity, and currents (Johnson et al., 2015). Floats capable of profiling to 4000 dbar and 6000 dbar have been designed and successfully deployed in regional pilot arrays (Roemmich et al., 2019), quantifying recent warming of bottom waters of Antarctic origin in the Southwest Pacific (Johnson et al., 2019) and Brazil Basins (Johnson et al., 2020), detecting temperature-salinity variability near Antarctic Bottom Water formation regions (Thomas et al., 2020; Foppert et al., 2021), and allowing detailed analyses of deep ocean circulation (Zilberman et al., 2020). A 6000-dbar profiling capability allows sampling from the ocean surface to the sea floor over 98% of the ocean area and 97% of the ocean volume, with the exceptions being a few very deep ocean trenches, and small portions of the ocean immediately above the deepest abyssal plains.
At the start of its descent phase, the Deep SOLO float model (Roemmich et al., 2019) pumps all of the oil that is required to target a 0.05 m s-1 descent rate when it reaches the bottom from its external bladder to its internal reservoir. Hence, it sinks most rapidly (order 0.25 m s-1) in the lightest waters near the surface, and steadily slows as it descends into denser waters. Since it does not make buoyancy adjustments during descent, unlike some other profiling floats, a flight model (e.g,, Cusack et al., 2017) is not needed to interpret variations around the background profiling rate as oceanic vertical velocities. The float controller, batteries, and buoyancy engine are inside mated 13” diameter glass hemispheres protected by plastic hard-hat housings, with the CTD (conductivity-temperature-depth) instrument in a smaller externally mounted titanium cylinder. Pressure samples are taken typically at between 5 and 50-dbar pressure intervals, depending on software settings, to monitor the descent rate. A 3-meter wire rope dangling from the float brakes its descent as the rope lays down on the bottom, preventing the instruments on the float from contacting the sea floor.
Initial examinations of vertical descent velocities from Deep SOLO float profiles readily revealed signatures of internal waves as oscillations around what would have otherwise been a steady deceleration with increasing pressure. The float-measured oscillations have average amplitudes of order 0.007 dbar s-1 and vertical wavelengths from ~400–1600 dbar in the water column below 1000 dbar. Here we explore regional variations in the amplitudes and wavelengths of these serendipitous observations of internal wave signals among and within the basins sampled by Deep Argo regional pilot arrays using Deep SOLO floats.